Fable pre-ban wouldn’t continue an old conversation about brain chemistry, sooo I guess they’re dusting off old tricks now that their new robot is offline.
Making Claude a Chemist
81–90 of 97 posts
Re: Making Claude a Chemist
#82Earlier quoted context omitted.
It would make more sense to have specialized versions of Claude instead of having a universal one ie BioClaude, ProgrammerClaude, NuclearClaude, CyberSecurityClaude, et al This would be both safer and less annoying to use.
That seems bad because it kills interdisciplinary science.
sure, but how else would you better guarantee safety? How else can we avoid annoying guardrails?
Besides, it's not like you can't have different agents with skills collaborating
Re: Making Claude a Chemist
#83inb4 someone calls Bessent to explain how this can be used in fentanyl production.
I know your comment is about the absurdity of such a thing, and maybe the stupidity of the administration, but for the readers at home:
Fentanyl is actually a very simple molecule. The way it's (and other drugs) production is controlled is by controlling and monitoring the precursor chemicals. So if someone goes out and buys a lot of proprionyl chloride, they will pop up on the DEA's list.
This is a much more sane thing to do than hemming and hawing about how dangerous the AI chatbot is and placing yourself into a position superiority over your users, pretending you know better than them, assuming that you must protect them from certain knowledge for their own good, etc, etc.
Re: Making Claude a Chemist
#84Earlier quoted context omitted.
That seems bad because it kills interdisciplinary science.
sure, but how else would you better guarantee safety? How else can we avoid annoying guardrails? Besides, it's not like you can't have different agents with skills collaborating
> how else would you better guarantee safety?
You wouldn't. This safety nonsense is overrated.
Re: Making Claude a Chemist
#85Let’s ban this before it gets too powerful !
I don't really want the average LLM being able to tell anyone exactly how to syntheize lethal nerve agents.
Don't worry, it is extremely likely that anyone who actually tries to synthesize such things will die trying before any serious damage is done.
Re: Making Claude a Chemist
#86Earlier quoted context omitted.
That seems bad because it kills interdisciplinary science.
sure, but how else would you better guarantee safety? How else can we avoid annoying guardrails? Besides, it's not like you can't have different agents with skills collaborating
Safety is a lie. It's always possible to divide a task into subtasks, to use abstractions, and the LLM will allow each of these individually.
For example, If you want it to write malware, break that down into a hundred high level steps, then implement each one individually, then stitch them together.
A dumber but uncensored agentic LLM will even do this for you to control the censored smarter LLM.
Re: Making Claude a Chemist
#87Re: Making Claude a Chemist
#88Let’s ban this before it gets too powerful !
I don't really want the average LLM being able to tell anyone exactly how to syntheize lethal nerve agents.
Aum Shin Rikyo didn't need an LLM to synthesize Sarin, and neither do you.
But when the police raid the hideout of the next Aum Shin Rikyo successor group and find evidence that they got the recipe from an LLM rather than traditional chemistry books, the resulting kneejerk legislative overreaction will cost us all some freedom.
That's not OK.
Re: Making Claude a Chemist
#89Earlier quoted context omitted.
> mining enough yellow cake and then purifying it is an industrial scale operation, out of reach unless you are a nation state or a transnational (or even national scale) energy and or minerals company. Might be hard to slip past the shareholders, but dark projects have flown under the annual reports of several large players.
Highly unlikely. Just the energy cost of running centrifuges, let alone manufacturing them is impossible to hide. It's staggering when you consider it. To put some more perspective, an LLM could show you how to make a Boeing 747 in detail, but the actual chance of someone making one with no tools and logistics would be 0. LLMs haven't changed the economic realities.
> Highly unlikely. Just the energy cost of running centrifuges, let alone manufacturing them is impossible to hide.
And yet that has happened twice already. Eg:
NUCLEAR ANXIETY: THE BLUNDERS; U.S. Blundered On Intelligence, Officials Admit
https://www.nytimes.com/1998/05/13/world/nuclear-anxiety-the...
There are numerous large companies with the resources and technical know how to match either Pakistan or India's efforts here, companies already experienced in yellowcake extraction and complex production processing equipment manufacture.
> To put some more perspective, an LLM could show you how to make a Boeing 747 in detail, but the actual chance of someone making one with no tools and logistics would be 0.
Pakistan, India, and many companies have resources, tools, logistics experience .. I have no idea how this sentence is meant to meaningfully refute those facts.
> LLMs haven't changed the economic realities.
I don't recall mentioning LLM's and I remind all that the bulk of nuclear weapons development happened before LLM's were a thing.
FWiW there are better actual arguments why it is unlikely (but not impossible) for a non state actor to put together a nuclear weapon .. the comment above is not one.
Re: Making Claude a Chemist
#90Why does chemistry require inference?